LSSBB Practice Questions

This is a primary practical application of the CLT. It states that the sampling distribution of the sample means will be approximately normal for large sample sizes (often n > 30), regardless of the population's distribution. This allows for the valid use of tests that assume normality of the means.

The fact that the sampling distribution of the mean is normal allows us to use z-scores or t-scores to calculate the margin of error and construct confidence intervals around the sample mean to estimate the true population mean.

In a 2^k full factorial design, the number of runs is 2 raised to the power of k, where k is the number of factors. For a 2^4 design, the number of runs is 222*2 = 16. The number of main effects that can be estimated is equal to the number of factors, which is 4 in this case.

This is a classic example of Poka-Yoke, or mistake-proofing. The software change makes it impossible for the error (using an outdated code) to occur. It's a prevention-type control that forces the correct action. Kanban is a signaling system for inventory control, 5S is a workplace organization methodology, and Kaizen is a philosophy of continuous improvement.

Rolled Throughput Yield is calculated by multiplying the First Time Yield (FTY) of each process step together. This gives the probability that a unit will pass through the entire process defect-free on the first attempt, without needing any rework. This method highlights the cumulative effect of defects and exposes the 'hidden factory' of rework loops that simple FTY calculations at each step might conceal.

The diagram visually indicates that the largest source of variation is temporal, or between batches. The link between 'Batch 1 Avg' and 'Batch 2 Avg' is labeled 'High Variation'. In contrast, the positional and cyclical components are labeled 'Low Variation'. Therefore, the team should prioritize investigating factors that change between batches, such as raw material lots, environmental conditions, or setup procedures.

A Multi-Vari Chart is the correct tool because it is specifically designed to display multiple sources of variation on a single graph. It can effectively show positional (within-piece), cyclical (piece-to-piece), and temporal (time-to-time) variation, making it ideal for pinpointing the largest contributor to overall process variability in this scenario. Box plots are good for comparing distributions but don't show all three types of variation simultaneously. Interaction plots are used after a DOE to visualize interactions, and a Main Effects Plot shows the effect of one factor at a time.

The most critical finding is that the number of distinct categories (ndc) is 3. An ndc of less than 5 indicates that the measurement system cannot adequately distinguish between different parts. This is a primary indicator of an unacceptable system, regardless of other metrics. While a %Tolerance of 25% is borderline acceptable (10-30% may be acceptable based on application), the low ndc value is a definitive failure point. The low Part-to-Part variation %Contribution (18%) also suggests the measurement error is consuming a large portion of the observed variation, reinforcing the conclusion that the system is inadequate.

A D-Optimal custom design is the most appropriate choice. This approach is specifically suited for situations with mixed-level factors (3, 3, and 2 levels) and a strict constraint on the number of runs (15). D-Optimal designs algorithmically select the most informative subset of possible runs to efficiently estimate the desired model terms (main effects, interactions, quadratic terms). A full factorial (3x3x2=18 runs) exceeds the budget. A Box-Behnken design requires factors to have 3 levels and is not directly applicable to the mixed-level scenario. A 2k-p fractional factorial is only for factors with 2 levels.

A control plan must clearly define what is being controlled (the characteristic) and what the acceptable range of performance is (specification limits).

The plan must detail how the characteristic will be measured, how often, and how many samples to take. This ensures consistent monitoring.

A critical component is the reaction or response plan. This pre-defines the steps to take when an out-of-control condition is detected, ensuring a swift and standardized response to problems.

This statement is true. A high Cpk (1.50) indicates excellent potential capability based on short-term, within-subgroup variation. However, a much lower Ppk (0.90) which uses long-term, overall variation, signifies that the process performance degrades over time. This discrepancy is a classic indicator of instability, special cause variation, or shifts and drifts between subgroups.